On 14 November 2013 09:26, Shmuel Metz (Seymour J.) <[email protected]> wrote: > at 08:40 PM, David Crayford <[email protected]> said: > >>Or patents! I notice IBM have quite a few wrt PLO. > > Could they be defensive patents? The only really effective way to > prevent someone else from patenting a technique that you're using is > to file first.
If you don't want to patent, but also don't want others to patent your invention, you can publish in a place that gets reasonable exposure, such as an academic journal in the field. Or if you're big enough (like IBM), you can publish your own Technical Disclosure Journal just for this purpose. Of course if you want the additional benefit of having your competition not know of your publication, then you need to publish in the most obscure journal in the most obscure language that you can away with while still claiming that it's prior art. IIRC there was a case in the 1980s or 90s where a US company published in Chinese in the Acta Chimica Sinica, which the USPTO then claimed wasn't likely enough to ever be seen by US researchers. Subsequently Microsoft and other software companies have done the same, but since then it's become more reasonable to expect that Chinese journals will be noticed. Meanwhile, IBM has discontinued its TDJ, perhaps for other reasons, but perhaps because they're being more aggressive with patents as a matter of policy. > Unless and until the USPTO gets serious about excluding > things that are obvious to a practitioner or are prior art, I expect > to continue to see a flood of defensive patents, and not just from > IBM. IBM has patented a number of new instructions straight out of the Principles of Operation, with no attempt to make the claims general. Presumably this is just to make life harder for would-be suppliers of competitive hardware. For example Move with Optional Specifications got US patent 20070277014, although the claims are just a particular application of the age-old notion of calling a subroutine with arguments and doing different things based on those arguments. I can't imagine it standing up to a challenge, but that costs a lot of money, and when you have hundreds of instructions to patent... > Note that even if the USPTO rejects a defensive patent as prior art, it will > have served its purpose. Indeed, and it may perhaps be cheaper than other avenues. Tony H. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
